一种自适应Retinex的航空影像阴影消除方法

摘要

In aerial images,there are many shadows covered by buildings,which not only influence the visibility of the images,but also the recognition of objects.It is necessary to compensate and remove the shadows.The Retinex is an image enhancement algorithm which can improve the brightness,contrast and sharpness of the images.A self-adaptive Retinex is proposed to overcome some limitations of the traditional Retinex.The self-adaptive Retinex can provide the parameters based on the average gray of the shadowed regions.Experimental results show this method can be effective.%航空影像中,由于高大建筑物的遮挡,常有阴影区域存在,这不仅会影响视觉效果,而且对建筑物的识别和提取带来困难,必须对其进行消除或者补偿.针对传统单尺度Retinex图像增强算法的不足,提出一种自适应单尺度Retinex算法.自适应单尺度Retinex算法首先对影像中阴影区域与非阴影区域作出概略区分,而后根据影像的灰度信息自适应计算Retinex算法的参数.实验证明,所提出的算法同经典的Retinex算法相比,无论是在增强效果还是处理时间上都具有一定的优越性.